Weekday Market Patterns

# Add this to the beginning of your load-market-data chunk

# Create a data directory if it doesn't exist
if (!dir.exists("data")) {
  dir.create("data", showWarnings = FALSE)
}

# Define constants for data processing
MAX_RETURN_VALUE <- 20  # Cap at ±20% for daily returns as a reasonable bound
MAX_REGIONAL_RETURN <- 5  # Cap at ±5% for regional averages
WINSORIZE_PERCENTILE <- 0.005  # Winsorize at 0.5% on each tail

# Create a robust cache file path system for storing data
cache_file <- file.path("data", "global_market_data_cache_5yr.rds")
cache_file_weekly <- file.path("data", paste0("global_market_data_cache_5yr_", format(Sys.Date(), "%Y_%W"), ".rds"))
cache_file_daily <- file.path("data", paste0("global_market_data_cache_5yr_", format(Sys.Date(), "%Y_%m_%d"), ".rds"))

# Function to safely read cache files
safe_read_cache <- function(file_path) {
  tryCatch({
    if (file.exists(file_path)) {
      message(paste("Attempting to load cache from:", file_path))
      data <- readRDS(file_path)
      
      # Basic validation
      if (is.data.frame(data) && nrow(data) > 0 && 
          all(c("date", "symbol", "close") %in% colnames(data))) {
        message(paste("Successfully loaded cache with", nrow(data), "rows from", file_path))
        return(data)
      } else {
        message("Cache data structure is invalid or empty")
      }
    }
    return(NULL)
  }, error = function(e) {
    message(paste("Error reading cache file:", e$message))
    return(NULL)
  })
}

# Function to safely save cache
safe_save_cache <- function(data, file_path) {
  tryCatch({
    # Create directory if needed
    dir.create(dirname(file_path), showWarnings = FALSE, recursive = TRUE)
    
    # Save data
    saveRDS(data, file_path)
    message(paste("Successfully saved cache to:", file_path))
    return(TRUE)
  }, error = function(e) {
    message(paste("Failed to save cache to:", file_path, "-", e$message))
    return(FALSE)
  })
}

market_data <- NULL

# Try to load from cache files in order of recency
if(is.null(market_data)) {
  market_data <- safe_read_cache(cache_file_daily)
}

if(is.null(market_data)) {
  market_data <- safe_read_cache(cache_file_weekly)
}

if(is.null(market_data)) {
  market_data <- safe_read_cache(cache_file)
}

# If no cache, new code for fetching data will run as defined in your existing logic

# A tibble: 10 × 2
   region          count
   <chr>           <int>
 1 Western Europe      7
 2 East Asia           5
 3 Middle East         5
 4 North America       5
 5 South Asia          5
 6 Africa              4
 7 Eastern Europe      4
 8 Latin America       4
 9 Northern Europe     3
10 Oceania             2

Global Patterns: Is There a “Best Day” for Markets?

The chart below shows average returns by weekday across all global markets. Green bars show better performance, while red bars show weaker days.

Markets included in this analysis: 🇦🇪 Abu Dhabi, 🇦🇷 MERVAL Argentina, 🇦🇺 S&P/ASX 200, 🇧🇷 Bovespa, 🇨🇦 S&P/TSX Composite, 🇨🇭 Swiss Market Index, 🇨🇱 S&P CLX IPSA, 🇨🇳 CSI 300, 🇩🇪 DAX 40, 🇩🇰 OMX Copenhagen, 🇪🇬 Egypt EGX 30, 🇪🇸 IBEX 35, 🇪🇺 EURO STOXX 50, 🇫🇷 CAC 40, 🇬🇧 FTSE 100, 🇭🇰 Hang Seng, 🇮🇩 Jakarta Composite, 🇮🇳 BSE SENSEX, 🇮🇳 Nifty 50, 🇮🇹 FTSE MIB, 🇯🇵 Nikkei 225, 🇰🇷 KOSPI, 🇲🇽 IPC Mexico, 🇲🇾 Malaysia KLCI, 🇳🇬 Nigeria SE, 🇳🇴 Oslo Exchange, 🇵🇱 Poland WIG, 🇷🇺 MOEX Russia, 🇸🇪 OMX Stockholm, 🇸🇬 Straits Times, 🇹🇷 BIST 30, 🇹🇼 Taiwan Weighted, 🇺🇸 Dow Jones, 🇺🇸 Nasdaq, 🇺🇸 Russell 2000, 🇺🇸 S&P 500, 🇿🇦 JSE Top 40

Regional Comparison: Do Weekday Effects Vary by Region?

Different regions around the world show different weekday patterns, which may be influenced by local trading behavior, timezone differences, and market structure.

Check Your Market’s Pattern

Different markets can show different patterns. Use the navigation arrows to explore weekday performance for each market:

Best and Worst Days by Region

The table below shows the best and worst performing weekdays for each region: